The challenge of marketing effectiveness in a cookieless world
Third-party cookies have been the mainstay of digital advertising, targeting and multi-touch attribution since its inception. However, the tech giants are responding to privacy legislation and concerns by removing the ability to track consumers across websites, apps and devices. Even with Google’s delay, this is a juggernaut of a trend that is unlikely to reverse in any meaningful way, as privacy regulations globally take hold and ‘walled gardens’ hold on to the value of the data they collate. While 2021 saw attribution at an individual level take a dive, the reality is that addressability and tracking has been in gradual decline for a number of years. And while there are alternative solutions for targeting, such as what’s on offer from walled gardens and contextual targeting, measurement is more tricky. Consequently, brands must grasp the nettle if they’re to see success in marketing effectiveness.
An opportunity to really get rid of data without understanding!
The easy availability of incredibly granular digital data generated an industry in analysing it to deliver instant returns. You can see why marketing teams increasingly focused on digital marketing over the last decade. And how multi-touch attribution became a mainstay of measurement. But often this has come with accompanying (and expensive) large-scale data projects that fail to keep pace as data availability changes. The result has been an over-emphasis on digital channels, which has come at the expense of the bigger picture, either in attention or value, with an understanding of the genuine influences on customer behaviour de-prioritized.
The focus flipped from outcome to method, and with it, less consideration for quality, repeatable, revenue generation in favour of hyper-personalisation and a fast buck. The tail has been wagging the dog. The reduction in our ability to measure at an individual level has had not only a lasting impact on measurement, but also on digital marketing costs. Many brands are fishing in the same decreasing pool or hampered by an inability to optimize as they did before. Plus the media landscape shifts and changes all the time, for example with ‘traditional’ channels increasingly moving into the digital space, and smart technology generating yet more signals. The challenge of understanding the impact of marketing and advertising activity has never been greater.
The good news is… attribution has never been the (only) answer!
Marketing effectiveness fundamentally depends upon our ability to understand how all channels, devices, opportunities to see and interactions work together to deliver outcomes, including those typically regarded as less measurable.
Marketing analytics seeks to understand the actions that drive brand affinity and incremental purchase, in other words, a purchase that would not have happened without that action impacting a consumer. It has long been understood that first and last click attribution are flawed as they take no account of the rest of the journey – though last click has still been the most used, for the most-part as multi-touch attribution (MTA) is so difficult to achieve. MTA works by understanding the touchpoints that consumers have interacted with and, using a mathematical approach, assigning their value or weight in the purchase – there are a number of different algorithms used in the space. But even multi-touch attribution has never effectively taken account of all channels, especially above the line, such as TV or out of home. This makes it incredibly difficult to be confident in how to compile and spend budget to best effect. On the one hand, granularity can understand the impact of micro changes to the tiniest of levers in a digital campaign. On the other, brand measures are typically intangible and based on ‘gut feel’. The resultant famine or feast in measurement confidence makes for a tough gig when explaining yourself to your CFO. Plus data comes in and out of availability all the time, so it’s important that measurement solutions can take account of that.
We need to be clear on the role of each data source within an overall framework, because ‘in the path’ doesn’t always mean impacting the path. That doesn’t mean the touchpoint is valueless, but it also doesn’t mean it is driving ROI! The need now to move from an (apparent) detailed understanding of performance and an unknown value associated with brand, to a new model that understands marketing at every stage of the funnel is a good thing, if an awkward about turn.
The trouble is, there is no new standard
Marketers are faced with a confusing array of options, each with their own sophisticated, privacy-conscious but incomplete solution that lacks interoperability across platforms. Moreover, some solutions have lost high-profile legal cases in recent times, with more legal wrangling to come. Fundamentally, there is no magic solution to replacing identifiers and chasing it is not the way forward. Far more important is ensuring the overall strategy and approach to measurement is right, which will then lead the way to the data solutions uniquely employed by each business. Consented first party data is the holy grail for any marketer – it’s deterministic, such as an email address or even good old-fashioned names and addresses, and allows for laser sharp targeting and granular measurement and analysis within owned platforms. But it’s not relevant or necessary for everyone!
Even if first party data is an option, its use is far from ubiquitous and gaining consent is a bit of a dating game, where the provider of the data expects something in return – both a benefit and the respect for their information. They are placing their trust in the brands they give this to and don’t expect that trust to be abused or they will rescind permission to use the data and potentially be lost forever. Which means a first party data strategy must also have the foundations to use it well. We have seen that the likes of Facebook, Apple and Google are in effect realizing the value of their own data via their walled gardens. Many other publishers are also creating similar environments in data clean rooms, with some working together in publisher networks to make their valuable first party data available to advertisers in an anonymous format that offers targeting and measurement opportunities. Better identity solutions that uphold the intention of privacy behind the changes to identifiers will undoubtedly emerge over time, but it is likely to be messy for the foreseeable future…
Remember that identity is not the only game in town!
Effective targeting and measurement can come from other quarters, such as contextual targeting and econometrics. So it also falls to marketers to demand more of data science, not only to plug these identity gaps, but to take a fresh look at measurement.